Multi - Modal Medical Image Fusion Algorithm Based On Discrete Wavelet Transform

Multimodal medical image fusion is the only possible way to integrate all relevant and complementary information from multiple modality images into single composite image, which is widely used in clinical application such as noninvasive diagnosis, image-guided radiotherapy, and treatment planning. Nowadays a lot of image fusion algorithms have been proposed, however, the study and analysis of medical image fusion is still a challenging area of research.in this paper, we presented a multiscale fusion of multimodal medical images in wavelet domain, the detail information are as follows.(1) Propose a novel Algorithm for multi-modality medical image fusion based on discrete wavelet transform. The quality of fused image lies on the fusion algorithm, the number of decomposition levels, and the fusion rule, to obtain better fusion performance, we focus on the research of the fusion rule. According to the characteristics of medical image, this paper designs new fusion rules for high-frequency and low-frequency coefficients on the basis of discrete wavelet transform, estimate different decomposition levels, and carry the consistency check on the fused images. Through comparing with the fusion results based on different fusion rules, different fusion algorithms, different decomposition levels, noise environment, and a clinical example, we can draw the conclusion that the proposed fusion rule is suitable for medical image fusion, which can efficiently preserve most information of source images and provide clear and detail parts. In addition, the combined subjective and objective evaluations of the proposed fusion method at multiple scales shows the effectiveness and goodness than other approaches even in the noisy environment, and achieves better performance in the clinical application.(2)Wavelet transform can’t decompose the high frequency signal, the wavelet packet transform is introduced into overcome the limitation. We adopt the same fusion rule that mentioned above, the comparative result between wavelet transform and wavelet packet transform is as follows. As the decomposition level is less, the wavelet packet transform can provide more abundant detail information, higher relevance with the source images, higher resolution, which indicates wavelet packet transform is superior to wavelet transform; much difference between the two fusion methods as the decomposition level is high, the fused image based on wavelet transform is with higher distortion and larger calculation, whereas the fused image becomes much better by using wavelet transform which has a clear advantage obviously. In conclusion, the wavelet packet transform is not necessarily more competitive than wavelet transform in the medical image fusion, it is advisable to select appropriate transform and decomposition level according to the actual demand.In conclusion, the proposed fusion rule can take high-frequency and low-frequency sub-images into account, which overcome the limitation of most fusion rules. The high-quality fused images will be obtained by using wavelet transform and wavelet pocket transform in basis of the proposed fusion rules, and which one will be selected depend on the clinical demand and the decomposition levels.